BENEDICT ANDERSON: Imagined Communities

Before proceeding to a discussion of the specific origins of nationalism, it may be useful to recapitulate the main propositions put forward thus far. Essentially, I have been arguing that the very possibility of imagining the nation only arose historically when, and where, three fundamental cultural conceptions, all of great antiquity, lost their axiomatic grip on men's minds. The first of these was the idea that a particular script-language offered privileged access to ontological truth, precisely because it was an inseparable part of that truth. It was this idea that called into being the great transcontinental sodalities of Christendom, the Ummah Islam, and the rest. Second was the belief that society was naturally organized around and under high centres—monarchs who were persons apart from other human beings and who ruled by some form of cosmological (divine) dispensation. Human loyalties were necessarily hierarchical and centripetal because the ruler, like the sacred script, was a node of access to being and inherent in it. Third was a conception of temporality in which cosmology and history were indistinguishable, the origins of the world and of men essentially identical. Combined, these ideas rooted human lives firmly in the very nature of things, giving certain meaning to the everyday fatalities of existence (above all death, loss, and servitude) and offering, in various ways, redemption from them.
The slow, uneven decline of these interlinked certainties, first in Western Europe, later elsewhere, under the impact of economic change, 'discoveries' (social and scientific), and the development of increasingly rapid communications, drove a harsh wedge between cosmology and history. No surprise then that the search was on, so to speak, for a new way of linking maternity, power and time meaningfully together. Nothing perhaps more precipitated this search, nor made it more fruitful, than print-capitalism, which made it possible for rapidly growing numbers of people to think about themselves, and to relate themselves to others, in profoundly new ways.
If the development of print-as-commodity is the key to the generation of wholly new ideas of simultaneity, still, we are simply at the point where communities of the type 'horizontal-secular, transverse-time' become possible. Why, within that type, did the nation become so popular? The factors involved are obviously complex and various. But a strong case can be made for the primacy of capitalism.
As already noted, at least 20,000,000 books had already been printed by 15001, signaling the onset of Benjamin's 'age of mechanical reproduction.' If manuscript knowledge was scarce and arcane lore, print knowledge lived by reproducibility and dissemination2. If, as Febvre and Martin believe, possibly as many as 200,000,000 volumes had been manufactured by 1600, it is no wonder that Francis Bacon believed that print had changed 'the appearance and state of the world."
One of the earlier forms of capitalist enterprise, book-publishing felt all of capitalism's restless search for markets. The early printers established branches all over Europe: 'in this way a veritable "international" of publishing houses, which ignored national [sic] frontiers, was created'4. And since the years 1500-1550 were a period of exceptional European prosperity, publishing shared in the general boom. 'More than at any other time' it was 'a great industry under the control of wealthy capitalists." Naturally, 'book-sellers were primarily concerned to make a profit and to sell their products, and consequently they sought out first and foremost those works which were of interest to the largest possible number of their contemporaries.'6
The initial market was literate Europe, a wide but thin stratum of Latin readers. Saturation of this market took about 150 years. The determinative fact about Latin—aside from its sacrality—was that it was a language of bilinguals. Relatively few were born to speak it and even fewer, one imagines, dreamed in it. In the sixteenth century the proportion of bilinguals within the total population of Europe was quite small; very likely no larger than the proportion in the world's population today, and—proletarian internationalism notwithstanding—in the centuries to come. Then and now the vast bulk of mankind is monoglot. The logic of capitalism thus meant that once the elite Latin market was saturated, the potentially huge markets represented by the monoglot masses would beckon. To be sure, the Counter-Reformation encouraged a temporary resurgence of Latin-publishing, but by the mid-seventeenth century the movement was in decay, and fervently Catholic libraries replete. Meantime, a Europe-wide shortage of money made printers think more and more of peddling cheap editions in the vernaculars.7
The revolutionary vernacularizing thrust of capitalism was given further impetus by three extraneous factors, two of which contributed directly to the rise of national consciousness. The first, and ultimately the least important, was a change in the character of Latin itself. Thanks to the labours of the Humanists in reviving the broad literature of pre-Christian antiquity and spreading it through the print-market, a new appreciation of the sophisticated stylistic achievements of the ancients was apparent among the trans-European intelligentsia. The Latin they now aspired to write became more and more Ciceronian, and, by the same token, increasingly removed from ecclesiastical and everyday life. In this way it acquired an esoteric quality quite different from that of Church Latin in mediaeval times. For the older Latin was not arcane because of its subject matter or style, but simply because it was written at all, i.e. because of its status as text. Now it became arcane because of what was written, because of the language-in-itself.
Second was the impact of the Reformation, which, at the same time, owed much of its success to print-capitalism. Before the age of print, Rome easily won every war against heresy in Western Europe because it always had better internal lines of communication than its challengers. But when in 1517 Martin Luther nailed his theses to the chapel-door in Wittenberg, they were printed up in German translation, and 'within 15 days [had been] seen in every part of the country.8 In the two decades 1520-1540 three times as many books were published in German as in the period 1500-1520, an astonishing transformation to which Luther was absolutely central. His works represented no less than one third of all German-language books sold between 1518 and 1525. Between 1522 and 1546, a total of 430 editions (whole or partial) of his Biblical translations appeared. 'We have here for the first time a truly mass readership and a popular literature within everybody's reach.9 In effect, Luther became the first best-selling author so known. Or, to put it another way, the first writer who could 'sell' his new books on the basis of his name.10
Where Luther led, others quickly followed, opening the colossal religious propaganda war that raged across Europe for the next century. In this titanic 'battle for men's minds', Protestantism was always fundamentally on the offensive, precisely because it knew how to make use of the expanding vernacular print-market being created by capitalism, while the Counter-Reformation defended the citadel of Latin. The emblem for this is the Vatican's Index Librorum Prohibitorum—to which there was no Protestant counterpart—a novel catalogue made necessary by the sheer volume of printed subversion. Nothing gives a better sense of this siege mentality than Francois I's panicked 1535 ban on the printing of any books in his realm—on pain of death by hanging! The reason for both the ban and its unenforceability was that by then his realm's eastern borders were ringed with Protestant states and cities producing a massive stream of smugglable print; To take Calvin's Geneva alone: between 1533 and 1540 only 42 editions were published there, but the numbers swelled to 527 between 1550 and 1564, by which latter date no less than 40 separate printing-presses were working overtime.11
The coalition between Protestantism and print-capitalism, exploiting cheap popular editions, quickly created large new reading publics—not least among merchants and women, who typically knew little or no Latin—and simultaneously mobilized them for politico-religious purposes. Inevitably, it was not merely the Church that was shaken to its core. The same earthquake produced Europe's first important non-dynastic, non-city states in the Dutch Republic and the Commonwealth of the Puritans. (Francois I's panic was as much political as religious.)
Third was the slow, geographically uneven, spread of particular vernaculars as instruments of administrative centralization by certain well-positioned would-be absolutist monarchs. Here it is useful to remember that the universality of Latin in mediaeval Western Europe never corresponded to a universal political system. The contrast with Imperial China, where the reach of the mandarin bureaucracy and of painted characters largely coincided, is instructive. In effect, the political fragmentation of Western Europe after the collapse of the Western Empire meant that no sovereign could monopolize Latin and make it his-and-only-his language-of-state, and thus Latin's religious authority never had a true political analogue.
The birth of administrative vernaculars predated both print and the religious upheaval of the sixteenth century, and must therefore be regarded (at least initially) as an independent factor in the erosion of the sacred imagined community. At the same time, nothing suggests that any deep-seated ideological, let alone proto-national, impulses underlay this vernacularization where it occurred. The case of 'England'—on the northwestern periphery of Latin Europe—is here especially enlightening. Prior to the Norman Conquest, the language of the court, literary and administrative, was Anglo-Saxon. For the next century and a half virtually all royal documents were composed in Latin. Between about 1200 and 1350 this state-Latin was superseded by Norman French. In the meantime, a slow fusion between this language of a foreign ruling class and the Anglo-Saxon of the subject population produced Early English. The fusion made it possible for the new language to take its turn, after 1362, as the language of the courts—and for the opening of Parliament. Wycliffe's vernacular manuscript Bible followed in 1382.12 It is essential to bear in mind that this sequence was a series of 'state,' not 'national,' languages; and that the state concerned covered at various times not only today's England and Wales, but also portions of Ireland, Scotland and France. Obviously, huge elements of the subject populations knew little or nothing of Latin, Norman French, or Early English.13 Not till almost a century after Early English's political enthronement was London's power swept out of 'France'.
On the Seine, a similar movement took place, if at a slower pace. As Bloch verily puts it, 'French, that is to say a language which, since it was regarded as merely a corrupt form of Latin, took several centuries to raise itself to literary dignity',14 only became the official language of the courts of justice in 1539, when Francois I issued the Edict of Villers-Cotterets.15 In other dynastic realms Latin survived much longer—under the Habsburgs well into the nineteenth century. In still others, 'foreign' vernaculars took over: in the eighteenth century the languages of the Romanov court were French and German.16
In every instance, the 'choice' of language appears as a gradual, unselfconscious, pragmatic, not to say haphazard development. As such, it was utterly different from the self-conscious language policies pursued by nineteenth-century dynasts confronted with the rise of hostile popular linguistic-nationalisms. [ ... ] One clear sign of the difference is that the old administrative languages were just that: languages used by and for officialdoms for their own inner convenience. There was no idea of systematically imposing the language on the dynasts' various subject populations.17 Nonetheless, the elevation of these vernaculars to the status of languages-of-power, where, in one sense, they were competitors with Latin (French in Paris, [Early] English in London), made its own contribution to the decline of the imagined community of Christendom.
At bottom, it is likely that the esotericization of Latin, the Reformation, and the haphazard development of administrative vernaculars are significant, in the present context, primarily in a negative sense—in their contributions to the dethronement of Latin and the erosion of the sacred community of Christendom. It is quite possible to conceive of the emergence of the new imagined national communities without any one, perhaps all, of them being present. What, in a positive sense, made the new communities imaginable was a half-fortuitous, but explosive, interaction between a system of production and productive relations (capitalism), a technology of communications (print), and the fatality of human linguistic diversity.18
The element of fatality is essential. For whatever superhuman feats capitalism was capable of, it ibund in death and languages two tenacious adversaries.19 Particular languages can die or be wiped out, but there was and is no possibility of man's general linguistic unification. Yet this mutual incomprehensibility was historically of only slight importance until capitalism and print created monoglot mass reading publics.
While it is essential to keep in mind an idea of fatality, in the sense of a general condition of irremediable linguistic diversity, it would be a mistake to equate this fatality with that common element in nationalist ideologies which stresses the primordial fatality of particular languages and their association with particular territorial units. The essential thing is the interplay between fatality, technology, and capitalism. In pre-print Europe, and, of course, elsewhere in the world, the diversity of spoken languages, those languages that for their speakers were (and are) the warp and woof of their lives, was immense; so immense, indeed, that had print-capitalism sought to exploit each potential oral vernacular market, it would have remained a capitalism of petty proportions. But these varied idiolects were capable of being assembled, within definite limits, into print-languages far fewer in number. The very arbitrariness of any system of signs for sounds facilitated the assembling process.20 (At the same time, the more ideographic the signs, the vaster the potential assembling zone. One can detect a sort of descending hierarchy here from algebra through Chinese and English, to the regular syllabaries of French or Indonesian.) Nothing served to 'assemble' related vernaculars more than capitalism, which, within the limits imposed by grammars and syntaxes, created mechanically-reproduced print-languages, capable of dissemination through the market.21
These print-languages laid the bases for national consciousnesses in three distinct ways. First and foremost, they created unified fields of exchange and communications below Latin and above the spoken vernaculars. Speakers of the huge variety of Frenches, Englishes, or Spanishes, who might find it difficult or even impossible to understand one another in conversation, became capable of comprehending one another via print and paper. In the process, they gradually became aware of the hundreds of thousands, even millions, of people in their particular language-field, and at the same time that only those hundreds of thousands, or millions, so belonged. These fellow-readers, to whom they were connected through print, formed, in their secular, particular, visible invisibility, the embryo of the nationally-imagined community.
Second, print-capitalism gave a new fixity to language, which in the long run helped to build that image of antiquity so central to the subjective idea of the nation. As Febvre and Martin remind us, the printed book kept a permanent form, capable of virtually infinite reproduction, temporally and spatially. It was no longer subject to the individualizing and 'unconsciously modernizing' habits of monastic scribes. Thus, while twelfth-century French differed markedly from that written by Villon in the fifteenth, the rate of change slowed decisively in the sixteenth. 'By the 17th century languages in Europe had generally assumed their modern forms.22 To put it another way, for now three centuries these stabilized print-languages have been gathering a darkening varnish; the words of our seventeenth-century forebears are accessible to us in a way that his twelfth-century ancestors were not to Villon.
Third, print-capitalism created languages-of-power of a kind different from the older administrative vernaculars. Certain dialects inevitably were 'closer' to each print-language and dominated their final forms. Their disadvantaged cousins, still assimilable to the emerging print-language, lost caste, above all because they were unsuccessful (or only relatively successful) in insisting on their own print-form. 'Northwestern German' became Platt Deutsch, a largely spoken, thus sub-standard German, because it was assimilable to print-German in a way that Bohemian spoken-Czech was not. High German, the King's English, and, later, Central Thai, were correspondingly elevated to a new politico-cultural eminence. (Hence the struggles in late-twentieth-century Europe for certain 'sub-'nationalities to change their subordinate status by breaking firmly into print—and radio.)
It remains only to emphasize that in their origins, the fixing of print-languages and the differentiation of status between them were largely unselfconscious processes resulting from the explosive interaction between capitalism, technology and human linguistic diversity. But as with so much else in the history of nationalism, once 'there,' they could become formal models to be imitated, and, where expedient, consciously exploited in a Machiavellian spirit. Today, the Thai government actively discourages attempts by foreign missionaries to provide its hill-tribe minorities with their own transcription-systems and to develop publications in their own languages: the same government is largely indifferent to what these minorities speak. The fate of the Turkic-speaking peoples in the zones incorporated into today's Turkey, Iran, Iraq, and the USSR is especially exemplary. A family of spoken languages, once everywhere assemblable, thus comprehensible, within an Arabic orthography, has lost that unity as a result of conscious manipulations. To heighten Turkish-Turkey's national consciousness at the expense of any wider Islamic identification, Atattirk imposed compulsory communications.23 The Soviet authorities followed suit, first with an anti-Islamic, anti-Persian compulsory romanization, then, in Stalin's 1930s, with a Russifying compulsory Cyrillicization.24
We can summarize the conclusions to be drawn from the argument [ ... ] by saying that the convergence of capitalism and print technology on the fatal diversity of human language created the possibility of a new form of imagined community, which in its basic morphology set the stage for the modem nation. The potential stretch of these communities was inherently limited, and, at the same time, bore none but the most fortuitous relationship to existing political boundaries (which were, on the whole, the highwater marks of dynastic expansionisms).
Yet. it is obvious that while today almost all modern self-conceived nations—and also nation-states—have 'national print-languages,' many of them have these languages in common, and in others only a tiny fraction of the population 'uses' the national language in conversation or on paper. The nation-states of Spanish America or those of the 'Anglo-Saxon family' are conspicuous examples of the first outcome; many ex-colonial states, particularly in Africa, of the second. In other words, the concrete formation of contemporary nation-states is by no means isomorphic with the determinate reach of particular print-languages. To account for the discontinuity-in-connectedness between print-languages, national consciousness, and nation-states, it is necessary to turn to the large cluster of new political entities that sprang up in the Western hemisphere between 1776 and 1838, all of which self-consciously defined themselves as nations, and, with the interesting exception of Brazil, as (non-dynastic) republics. For not only were they historically the first such states to emerge on the world stage, and therefore inevitably provided the first real models of what such states should 'look like,' but their numbers and contemporary births offer fruitful ground for comparative enquiry.

[Imagined Communities (Verso: London, 1991), 36-46.1